Vulkan: Required LL2 extension prevents automatic fallback to legacy Reflex on older drivers (RTX 2080 Ti, 537.42)

Aperta
#132 0 commenti 0 reazioni 0 assegnatari Vedi su GitHub

Nessuno ha ancora preso questa issue.

Valutazione

Difficoltà
3/5
Tempo stimato
1-2 giorni
Idoneità per principianti
64/100
Tipo di issue
Bug
Chiarezza
Specificata chiaramente
Stato di attività
Attiva
Stack tecnologico
c
Ambito
backend

Direzione di ricerca

Read source/plugins/sl.reflex/reflexEntry.cpp and trace how the Turing+ heuristic adds VK_NV_low_latency2 to required_extensions, then compare it with optional_extensions filtering and the existing fallback from CreateVkNvLowLatency2 to CreateNvLowLatencyVk. Done means Vulkan device creation succeeds when LL2 is unavailable but the legacy extension is supported, and the legacy Reflex backend is selected.

Scritto dal modello di indicizzazione a partire dal testo della issue.

Descrizione

Streamline 2.14.16 fails Vulkan device creation on older NVIDIA drivers instead of reaching its existing automatic fallback to the legacy Reflex backend.

An affected configuration is:

  • GPU: GeForce RTX 2080 Ti
  • Driver: 537.42
  • Graphics API: Vulkan
  • Streamline: 2.14.16

This configuration supports the original VK_NV_low_latency path but does not expose VK_NV_low_latency2.

In [reflexEntry.cpp](https://github.com/NVIDIA-RTX/Streamline/blob/v2.14.16/source/plugins/sl.reflex/reflexEntry.cpp), the Turing+ architecture heuristic adds VK_NV_low_latency2 to the required device extension list. Consequently, the interposer returns VK_ERROR_EXTENSION_NOT_PRESENT during device creation:

Required device extension VK_NV_low_latency2 unsupported!

Streamline already implements an automatic fallback from CreateVkNvLowLatency2 to CreateNvLowLatencyVk. However, this fallback runs after device creation, so the required-extension check prevents it from ever being reached on this configuration.

The expected behavior is to create the device successfully and automatically use the original Reflex backend when LL2 is unavailable and the legacy backend is supported.

Could VK_NV_low_latency2 be consistently declared under optional_extensions, including on Turing+ GPUs? The interposer already filters optional extensions against the selected physical device’s capabilities. This would preserve LL2 on supported drivers while allowing the existing fallback to work on older drivers.

GPU architecture alone should not make LL2 mandatory when a compatible legacy Reflex path is available.

Lingua principale
C
Stelle
1.4k
Fork
174
Metriche di merge delle PR
Nessuna PR unita negli ultimi 30g

Preparare l'ambiente

Come iniziare

  1. Leggi tutta la issue e poi la guida ai contributi del progetto.
  2. Commenta sulla issue per dire che te ne occupi tu — evita che due persone facciano lo stesso lavoro.
  3. Fai un fork del repository e lavora su un branch.
  4. Apri una pull request che faccia riferimento al numero della issue.

Altre issue di NVIDIA-RTX/Streamline

Tutte le issue di NVIDIA-RTX/Streamline

Issue simili

Altre issue su C

Ricevi le nuove issue nella tua casella

Un breve riepilogo di issue GitHub adatte ai principianti.